Time-resolved fluorescence spectra of three amino-substituted coumarin dyes have been recorded in methanol and dimethyl sulfoxide using the fluorescence upconversion technique with an apparatus response function of ≈200 fs fwhm. The three fluorinated coumarins are the 7-amino-4-trifluoromethylcoumarin (C151), the 7-diethylamino-4-trifluoromethylcoumarin (C35), and the rigidified aminocoumarin with a julolidine structure (C153). The dynamic Stokes shifts are found to be dominated by an ultrafast component with a characteristic time shorter than the present time resolution of ≈50 fs. The dynamic Stokes shifts are compared to estimations based on a “Kamlet and Taft” analysis of steady-state data in 20 solvents. It is found that the ultrafast component can be assigned mainly to intramolecular relaxation. The influences of photoinduced changes of solute−solvent hydrogen bonds on the observed spectral shifts are discussed. The breaking of hydrogen bonds at the amino group is very fast in both solvents and embedded in the ultrafast solvent inertial relaxation, while the reformation of hydrogen bonds at the carbonyl group is believed to occur on the 10−20 ps time scale in the hydrogen bond donating (HBD) solvent methanol. However, it is impossible to unambiguously correlate a particular experimental time constant with the breaking or the formation of a hydrogen bond.

Fluorescence anisotropy decays of the 7-diethylamino-4-methylcoumarin C1 in various polar solvents of different viscosities and hydrogen bond donor/acceptor character have been recorded by means of the fluorescence upconversion and time-correlated single photon counting techniques. The resulting characteristic times for the rotational diffusion fall into two classes with regards to the viscosity-dependency: n-alcohols and "other" solvents. This deviation from the simple Stokes-Einstein-Debye model may be interpreted in terms of rotation of the coumarin molecule under two different hydrodynamic boundary-conditions ("stick" or "slip") in the two solvent classes. Possible explanations for this behaviour are discussed, and in particular solvent attachment and additional dielectric friction. Both these phenomena may in fact, under certain conditions, explain our findings. Our opinion, however, is that the dielectric friction model offers a more realistic picture of the additional rotational friction experienced by C1 in n-alcohols.